CLASH OF CULTURES: MUSLIMS AND CHRISTIANS IN THE

DISCUSSION PAPER SERIES

ABCD

www.cepr.org

Available online at: www.cepr.org/pubs/dps/DP5910.asp www.ssrn.com/xxx/xxx/xxx

No. 5910

CLASH OF CULTURES: MUSLIMS AND CHRISTIANS IN THE ETHNOSIZING PROCESS

Amelie Constant, Liliya Gataullina, Klaus F Zimmermann and Laura Zimmermann

LABOUR ECONOMICS

ISSN 0265-8003

CLASH OF CULTURES: MUSLIMS AND CHRISTIANS IN THE ETHNOSIZING PROCESS

Amelie Constant, IZA, Bonn, DIW Berlin and Georgetown University Liliya Gataullina, IZA, Bonn

Klaus F Zimmermann, Bonn University, IZA Bonn, DIW Berlin and CEPR Laura Zimmermann, University of Oxford and IZA

Discussion Paper No. 5910

October 2006

Centre for Economic Policy Research 90–98 Goswell Rd, London EC1V 7RR, UK

Tel: (44 20) 7878 2900, Fax: (44 20) 7878 2999 Email: [email protected], Website: www.cepr.org

This Discussion Paper is issued under the auspices of the Centre’s research programme in LABOUR ECONOMICS. Any opinions expressed here are those of the author(s) and not those of the Centre for Economic Policy Research. Research disseminated by CEPR may include views on policy, but the Centre itself takes no institutional policy positions.

The Centre for Economic Policy Research was established in 1983 as a private educational charity, to promote independent analysis and public discussion of open economies and the relations among them. It is pluralist and non-partisan, bringing economic research to bear on the analysis of medium- and long-run policy questions. Institutional (core) finance for the Centre has been provided through major grants from the Economic and Social Research Council, under which an ESRC Resource Centre operates within CEPR; the Esmée Fairbairn Charitable Trust; and the Bank of England. These organizations do not give prior review to the Centre’s publications, nor do they necessarily endorse the views expressed therein.

These Discussion Papers often represent preliminary or incomplete work, circulated to encourage discussion and comment. Citation and use of such a paper should take account of its provisional character.

Copyright: Amelie Constant, Liliya Gataullina, Klaus F Zimmermann and Laura Zimmermann

CEPR Discussion Paper No. 5910

October 2006

ABSTRACT

Clash of Cultures: Muslims and Christians in the Ethnosizing Process*

The paper explores the evolution of ethnic identities of two important and distinct immigrant religious groups. Using data from Germany, a large European country with many immigrants, we study the adaptation processes of Muslims and Christians. Individual data on language, culture, societal interactions, history of migration and ethnic self-identification are used to compose linear measures of the process of cultural adaptation. Two-dimensional variants measure integration, assimilation, separation and marginalization. Christians adapt more easily to the German society than Muslims. Immigrants with schooling in the home country and with older age at entry as well as female Muslims remain stronger attached to the country of origin. Female Muslims integrate and assimilate less and separate more than Muslim men, while there is no difference between male and female Christians. Christians who were young at entry are best integrated or assimilated, exhibiting lower separation and marginalization in the later years, while for Muslims a similar pattern is observed only for assimilation and separation. Christian immigrants with college or higher education in the home country integrate well, but Muslims do not. For both religious groups, school education in the home country leads to slower assimilation and causes more separation than no education at home. While school education has no impact on integration efforts for Muslim, it affects similar attempts of Christians negatively.

JEL Classification: F22, J15, J16, Z10 and Z12 Keywords: ethnic exclusion, ethnic identity, ethnicity, migrant assimilation, migrant integration and religion

Amelie Constant IZA Schaumburg-Lippe-Str. 7 / 9 D-53113 Bonn GERMANY Email: [email protected] For further Discussion Papers by this author see: www.cepr.org/pubs/new-dps/dplist.asp?authorid=159847

Liliya Gataullina IZA Schaumburg-Lippe-Str. 7 / 9 D-53113 Bonn GERMANY Email: [email protected] For further Discussion Papers by this author see: www.cepr.org/pubs/new-dps/dplist.asp?authorid=164433

Klaus F Zimmermann Director of IZA Schaumburg-Lippe-Str. 7 / 9 D - 53113 Bonn GERMANY Email: [email protected] For further Discussion Papers by this author see: www.cepr.org/pubs/new-dps/dplist.asp?authorid=102751

Laura Zimmermann Mansfield College Oxford OX 1 3TF UK Email: [email protected] For further Discussion Papers by this author see: www.cepr.org/pubs/new-dps/dplist.asp?authorid=165501

* Financial support from Volkswagen Foundation for the IZA project on “The Economics and Persistence of Migrant Ethnicity” is gratefully acknowledged. We wish to thank the IZA Volkswagen Ethnicity Research Team for valuable comments on earlier drafts.

Submitted 28 September 2006

1

1. Introduction

Religion and its role in the immigrants’ adaptation process to the host country are important issues that

have remained dormant until recently. After 9/11, these questions have strongly emerged in all

immigrant countries. In Germany, where almost every fifth person has an emigrational background

(Statistisches Bundesamt 2006), this issue is important. The number of people with religious beliefs

that are very different from the common religious beliefs of Germans and that often contradict the

German secular tradition is increasing, turning those people into major minority groups that face the

problems and bear the consequences of clashing cultures.

The scarce research undertaken so far has concentrated on ethnicity and often treated religion

just as an extension of this concept (Baumann 2000; Raj 2000; Kong 2001). Although religion and its

impact on the formation of the ethnic identity of immigrants has long been a relatively neglected field,

the rise of religiosity in multicultural societies has raised questions about the impact of religion on the

socio-economic integration process of migrants. This is especially interesting since religion seems to

shape individual traits like work ethics or trust with ensuing ramifications in the labour market (Guiso,

Sapienza and Zingales 2006; McCleary and Barro 2006). Studying the behaviour of immigrants of

different religions in a large immigrant European country like Germany - where 3.3 million Muslims

(or 4% of the total population) and 26.2 million Catholics (or 31.8% of the total population) among

54.3 million Christians (or 65.8% of the total population) live together (Religionswissenschaftlicher

Medien-Informationsdienst e. V., REMID, 2006) is therefore warranted.

The comparison of two religious groups, Muslims and Catholic Christians, is particularly

interesting. Muslims today form important minority groups in many European countries (as they do in

North America and elsewhere), and may often find it difficult to adjust to these host countries that

have different religious or secular traditions. In their analysis on the adaptation process of Muslim

workers in Japan, Onishi and Murphy-Shigematsu (2003) have documented these problems and

2

indicated three major ways for Muslims to position themselves in the Japanese society: assimilation,

which allows abolishing many practical problems of being different in Japan’s homogenous society;

marginalization that demonstrates no acceptance in either Japanese or native Muslim culture; and

separation from the Japanese society by an increased reliance on Islam. Other research reports that

while being Muslim in a multicultural society remains an important character trait it may also be one

of the key characteristics which make identifying with the host country possible, because religion is

not strongly bound to a particular place or country of origin (Nagata 1974; Jacobsen 1997; Saeed,

Blain and Forbes 1999; Kurien 2001). How Muslims adjust to life in Germany therefore is an

important and challenging question.

The adjustment process of Christians and especially Catholics is also of great interest. As

Christianity is still the main religion in Germany (although there is separation of church and state),

with Catholicism dominating many states of the country, Christian immigrants should find it easier to

adapt to life in Germany. However, research in the United States in the 1960s and 1970s shows that

Catholics acculturate rather quickly but they never really assimilate, keeping their own networks

(Parenti 1967; Alba 1976). Similar to the research on Muslims, the adjustment prediction of Christian

immigrants, especially Catholics, is not clear-cut. This is why the evaluation of how Christians and

especially Catholics adjust to the new situation in Germany is worthwhile.

In this paper we analyse the process of cultural adjustment of Muslim and Christian

immigrants in Germany. Our analysis employs two different approaches to modelling the concept of

cultural adaptation, or as we call it “ethnic identity.” Section 2 presents the measurement and data, and

lays out the hypotheses. Section 3 discusses the empirical results, and Section 4 draws conclusions.

3

2. Measurement, Data and Hypotheses

Ethnic identity is acknowledged to be a complex concept (Chávez and Guido-DiBrito 1999; Nazroo

and Karlsen 2003), which requires special attention for modelling. Following Constant, Gataullina and

Zimmermann (2006), we consider two approaches to measure ethnic identity: a one- and a two-

dimensional model. The one-dimensional model studies ethnic identity from the perspective that

attachment to the country of origin and attachment to the host country are mutually exclusive. That is,

a stronger attachment to the host country must mean a weaker connection to the country of origin and

vice versa; every combination of the two concepts is possible as long as both attachments sum up to

one. To measure the intensity of a person’s ethnic identity, we define the word ‘ethnosize’ as having

strong feelings for the country of origin, and construct the so-called one-dimensional ethnosizer from

information on five factor groups: language, culture, societal interaction, history of migration and

ethnic self-identification. These five variables are formed out of the mean value of answers given to all

the questions in the survey asked for each subgroup, ranging between zero (most commitment to

Germany) and one (most commitment to the country of origin) (see Constant, Gataullina and

Zimmermann 2006 for more details). The one-dimensional ethnosizer is the mean value of all five

factor groups, and ranges between zero and one as well.

While this approach is convenient, it is likely to overestimate immigrants’ attachment to the

host country as it automatically assumes that adaptation to the host country results in weaker

connections to the home country. Yet, as has been pointed out in the literature, ethnic identity can take

more complex forms and result in several permutations. Although adapting to the host country, people

today have various possibilities to keep strong ties with their home country (Haller and Landolt 2005).

Intensifying connections with the home country therefore does not necessarily mean cutting ties with

the country of origin and vice versa. A more realistic scenario is that the attachments to the host and

4

the home country are not always mutually exclusive, and there is room for more complex

combinations.

There are four possible ways of adapting one’s ethnic identity under the two-dimensional

ethnosizer: assimilation, that is complete adjustment to or absorption by the host country with cutting

all ties to the home country; integration, that is adjustment to the host country with simultaneous

retention of ties to the home country; separation, that is not adjusting to the host country but

withholding strong connections to the home country; and marginalization, that is having only weak

attachments to both the host and the home country (Berry 1980; Berry et al. 1989; de Domanico,

Crawford and de Wolfe 1994). The two-dimensional concept of the ethnosizer uses this classification

into the four states of integration, assimilation, separation and marginalization. For each state, we

generate a dependent variable that includes variables of all the five factor groups - language, culture,

societal interaction, migration history and ethnic self-identification - with paired information on

attachments to the host and the home country. As people can, for example, be integrated in one factor

group and separated in another, each dependent variable ranges between zero and five, measuring how

often a respondent is identified as integrated, assimilated, separated or marginalized in each of the five

factor groups. This approach allows the comparison among respondents as more or less integrated by

calculating, for instance, the number of categories in which a respondent was classified as integrated.

The same is possible for the other three measures of ethnic identity. For each immigrant, the

classifications add up to five.

The dataset used in the empirical analysis is the nationally representative German Socio-

economic Panel (GSOEP, SOEP Group 2001), conducted annually since 1984. Our sample contains

430 Muslim and 724 Christian and non-denominational first generation immigrants, and the base year

of observation chosen is 2001.

5

In analyzing ethnic identity econometrically, we use OLS regressions. The dependent variables

measuring both the one- and two-dimensional ethnosizers are: language, culture, societal interaction,

history of migration, ethnic self-identification, integration, assimilation, separation, and

marginalization (see above). As explanatory variables we use essential pre-migration characteristics

such as: age at entry (allowing for non linear effects) and education in the home country. The latter is a

categorical variable with no education being the reference category. We also control for gender and

age (allowing for non linear effects). In the analysis on Muslims, Turks are taken as the reference

group with dummy variables for the ex-Yugoslavian ethnicities and other ethnicities (including

Greeks). In the analysis on Christians, Catholics and Spaniards are taken as reference groups with

dummy variables for other Christian and non-denominational people for religion, and ex-Yugoslavs,

Greeks, Italians, and others for ethnicity.

We expect to find support for the following hypotheses: Starting from the considerations in the

introduction, we expect that Muslims remain more ethnic in the sense that they hold on to their culture,

and integrate and assimilate less and separate and marginalize more than Christians who can take

advantage of the fact that they have the same religion as about two thirds of the native population.

Education in the home country can possibly have a negative impact on adapting to Germany because

this means having been longer in one’s country of origin, having more knowledge and appreciation

about its society and traditions, and being educated and inculcated in a certain way. On the other hand,

though, more educated people may be more open-minded, and one of the reasons they migrate is to

find more educated people like them and broaden their horizons. Hence, they should integrate better in

Germany. Consequently, the net effect remains an empirical issue.

With regard to gender differences, we conjecture that males integrate and assimilate more

easily than females; this is especially expected for Muslim immigrants. If Muslim females in Germany

are more likely to stay home, while men go out and work, they have fewer interactions with the host

6

society, and therefore, integrate less easily because they are shielded at home or are not exposed in the

host society. The age at which an immigrant enters the host country is crucial in shaping the adaptation

process. We expect that immigrants who arrive at a younger age are more likely to amalgamate with

the host culture and be indistinguishable from natives in the long run. We control for the quadratic

term because the rate at which the adaptation process takes place varies.

3. Empirical Results

Table 1 presents the results of the OLS regression for Muslim immigrants using the linear ethnosizer

approach. As the last column shows, age at entry is a significant determinant of the ethnosizer, and the

older immigrants are when migrating to Germany, the more they remain connected to their country of

origin. Females and those immigrants with complete or incomplete schooling in the home country also

remain more ethnosized than men and those Muslims with no education in the home country.

However, ex-Yugoslav Muslims and Muslims of other ethnicities are less ethnosized and more

attached to Germany than Turkish Muslims.

Table 1 about here

Interesting are also the more detailed findings on the five components of the ethnosizer.

Column one shows that the results for the factor group language almost exactly resemble those for the

ethnosizer. Only age at entry squared is significant in addition to the other variables, showing that the

relationship between language and age at entry is not linear. In general, however, the determinants of

what makes one remain ethnic in relation to language - that is, what makes one more comfortable with

the language of the country of origin - are the same as for the ethnosizer. Column two presents the

results for the factor group culture. Here, too, what makes an immigrant culturally ethnic, meaning

what makes one connected to their country of origin in cultural respects, is mainly the same as for the

ethnosizer. Yet age is also significant indicating a strong non-linear relationship between age and

7

culture; the older one is, the more one is attached to the German culture and this occurs at a decreasing

rate.

The picture for societal interaction in column three is somewhat different. Age again plays an

important role; the older one is the more one interacts socially with non-Germans at a decreasing rate,

since the cubic term is small. Incomplete and complete schooling in the home country, as well as

vocational education at home all lead to less interaction with Germans compared with people without

any education in the home country. Muslim members of other ethnicities interact more with Germans

than Turkish or ex-Yugoslav Muslims. Column four shows the results for migration history. There are

some weakly significant age effects that are highly non-linear indicating a declining attachment to the

home country at younger ages. Other ethnicities are stronger connected to their country of origin than

Turks, whereas ex-Yugoslavs do not differ from Turks in this respect. In column five, we see that

Muslim females and those with complete or incomplete school education in the home country self-

identify as more ethnic than men and those with no education in the home country, whereas ex-

Yugoslav Muslims feel more German than Turkish Muslims or those of other ethnicities.

In general, we see that although the picture is more complex when looking at the components

than when just concentrating on the ethnosizer there are some general trends. Out of all Muslims,

females and those with some kind of schooling in the home country remain strongly ethnosized. Ex-

Yugoslav Muslims and those of other ethnicities usually are less ethnic than the Turks (the reference

group). A higher age at entry makes one more ethnosized with some discounting, and age has a

positive effect on attachment to Germany for culture and migration, but a negative effect on societal

interaction with Germans.

Table 2 presents the results on the two-dimensional ethnosizer model for Muslim immigrants.

Age at entry is strongly significant for assimilation and separation. As expected, the older immigrants

are when they enter Germany, the less likely they are to assimilate albeit at a decreasing rate. In

8

contrast, for each year older an immigrant arrives, the more likely he or she is to be in the separation

state at a decreasing rate. It is interesting that Muslim females are less likely to integrate or assimilate

and more likely to separate than Muslim men. This could be because Muslim females are more likely

to stay home shielded from the outside world and the influences of the host society. Whereas college

or higher education has no impact, incomplete schooling in the home country makes immigrants

assimilate less and separate and marginalize more. Muslims with vocational education in the home

country assimilate less, and Muslims with complete school education in the home country assimilate

less and separate more than those with no education in the home country. This implies that for

Muslims education in the home country is detrimental; their adjustment to the host country is easier

with no education at all from home. One possible explanation is that those Muslims have more likely

some German education. Compared to Turks, ex-Yugoslav Muslims and those of other ethnicities are

less likely to separate, while those of other ethnicities are more likely to assimilate.

Table 2 about here

Overall, Table 2 provides a similar picture to the one for the one-dimensional ethnosizer.

Muslim females cling stronger to their country of origin than males, as do those with some kind of

school education or vocational education in comparison to those without any education in the home

country. Ex-Yugoslavian Muslims and those of other ethnicities are less ethnic and therefore separate

less than Turkish Muslims. Immigrants who enter at an older age are more likely to remain ethnosized.

We now turn to the results on the ethnic identity of Christians, paying special attention to

Catholics. Table 3 shows the results for Christian and non-denominational immigrants, with Catholics

as the reference group, in the one-dimensional ethnosizer model. As we see, in all OLS regressions

Catholics do not differ from other Christians and non-denominational immigrants, and there are no

differences at all between males and females. The result for the ethnosizer in the last column shows

that, as expected, older immigrants upon arrival are more likely to remain ethnosized. Incomplete or

9

complete schooling in the home country make an immigrant more ethnosized than other immigrants

who have no education in the home country. Turkish, ex-Yugoslavian, Greek and Italian Catholics all

are more connected to their country of origin than the reference group of Spanish Catholics.

Table 3 about here

A more complex picture emerges when we look at the components of the ethnosizer in

columns one to five. In column one, we present the results for language. Again, what makes an

immigrant linguistically ethnosized is similar to what makes him or her ethnosized in general. There is

only one additional significant impact for those Catholics with college or higher education in the home

country, who are less linguistically ethnic than those with no education in the home country. In

column two, we report the results for culture. While most of the variables that affect the general

ethnosizer have also a significant impact on the cultural aspect of ethnic identity, there is an additional

non-linear relationship between age and culture. For societal interaction in column three, there is an

additional impact on those individuals of other ethnicities. These people are less ethnic, that is, they

have more contact with Germans, than the reference group of Spanish Catholics.

Column four shows the results on migration history. As age cubed and age at entry squared are

significant, we observe a non-linear relationship between age and migration history. However, both

estimated effect parameters are fairly small and only significant at the margin. Ex-Yugoslavs, Greeks

and those individuals of other ethnicities remain more connected to their country of origin than

Spanish Catholics. Where ethnic self-identification, as presented in column five is concerned, the older

one is upon entering the host country, the more ethnosized one remains. The same is true for those

immigrants with incomplete school education in the home country, and for Turks, Greeks and Italians

who all self-identify as more connected to their countries of origin than the reference group.

In general, Table 3 exhibits an interesting picture. For the Christian and non-denominational

immigrant sample, there is no difference between males and females or between Catholics, other

10

Christians and non-denominational immigrants in how they form their ethnic identity. Having some

kind of schooling in the home country makes immigrants more ethnic, the only exception being

college or higher education at home in the factor group language. Turkish, ex-Yugoslav, Greek, and

Italian Catholics are all more ethnosized than Spanish Catholics. Definitely, age at entry plays a

catalytic role in the ethnosizer with older people at arrival holding on to their ethnicity the most.

Table 4 presents the results for Christian and non-denominational immigrants in the two-

dimensional model. The older Catholics and the other members in this subgroup are, the less they

separate and the more they marginalize; both are non-linear relationships. Non-linear relationships also

exist for age at entry and the dependent variables assimilation and marginalization. The older an

immigrant is when entering Germany, the less assimilated or integrated and the more marginalized or

separated he or she is. The estimated effects on assimilation and marginalization are weaker with

rising age at entry. Again, we find no difference between males and females, or between Catholics and

other Christians or non-denominational people. While those with college or higher education at home

integrate better, those with complete and incomplete schooling in the home country integrate and

assimilate less and separate more than those with no education in the home country. Turkish, ex-

Yugoslav, Greek and Italian Christians or non-denominational immigrants assimilate less and separate

more, while Greeks and Italians integrate less than Spanish Catholics.

Table 4 about here

Overall, Table 4 depicts similar results for the one-dimensional ethnosizer: There is no

difference between males and females or between Catholics, other Christians and non-denominational

people in their ethnic identity formation. Having some schooling in the home country renders

immigrants more connected to their country of origin; the only exception being college or higher

education in the home country where people integrate better. Turkish, ex-Yugoslavian, Greek and

11

Italian Christians all remain more ethnosized than Spanish Catholics. Lastly, the older an immigrant is

when entering the country, the more ethnic he or she remains controlling for all other characteristics.

4. Summary and Conclusions

Comparing the behaviour of Muslim and Christian immigrants with regards to the formation of their

ethnic identity, the results of this study exhibit intriguing differences and similarities. While there are

no significant differences between Christian men and women, Muslim females integrate and assimilate

less and separate more than Muslim men. This verifies our hypothesis about gender differences for the

Muslim subsample. Age at entry also plays an important role. Older immigrants remain strongly

ethnosized. For Christians, being young when entering Germany has a positive effect on integration

and assimilation and makes separation and marginalization less likely. For Muslims, the older age at

entry only affects assimilation and separation in a negative and positive way respectively.

In general, immigrants with some level of schooling in the home country remain more ethnic,

and assimilate less and separate more than people with no education in the home country. The only

exception is Christian immigrants with college or higher education in the home country, who integrate

well. However, this is not the case for Muslims. Christians with school education in the home country,

on the other hand, integrate less than those with no education in the home country, while there is no

effect of school education in the home country on the integration of Muslims. These results support

our hypothesis that education in the home country has a negative impact on adapting to the host

country.

Overall, we see that Christian immigrants adapt more easily to the new situation in the host

country than Muslims. This supports the hypothesis that Muslims have problems forming an ethnic

identity in a country where the dominant religion is not Islam, whereas Christians take advantage of

the common religion. Yet, these findings also raise the question of what can be done in order to

12

improve the adaptation process of Muslim immigrants. As Muslims already form important minority

groups in Germany and in other European societies, integrating these people is a necessity in order to

prevent the rise of parallel societies and to ensure the best well-being possible for all groups.

These results are especially pertinent and enlightening in an international context. Onishi and

Murphy-Shigematsu (2003), e.g., found that no Muslim workers adjusted to their new life in Japan

through integration. While this may be either due to the small number of observations or to the still

relatively homogenous Japanese society, it is noteworthy that integration does not seem to be a

dominant strategy in Japan. On the other hand, in our analysis there does not seem to be much

evidence on the phenomenon of increased religiosity and simultaneous identification with the host

country that could foster the integration of Muslims, as observed in the UK (Saeed, Blain and Forbes

1999). For Muslim immigrants living in Germany, the multicultural element that would permit them to

form strong ethnic ties to both their home and their host country still seems to be lacking. This links

well with Abdullah’s (1995) observation that whereas Muslims in Germany may often disengage from

their religion they do not build up a strong attachment to replace it; the only way out is for them to

adapt to Germany while at the same time they maintain their religious identity.

The results for Christians, and especially for Catholics, seem to be in line with a decreasing

research interest in this group; the topic also seems to have become a non-issue in the public debate.

At least in Germany, Catholic immigrants adapt fairly well. Even if only few Catholics assimilate

completely, claiming that this is a problem seems to be an outdated opinion in our modern world

where the chief aim should be integration and a peaceful symbiosis of cultures and mores so all

inhabitants of the country can be the best they can be.

13

References

ALBA, RICHARD D. 1976 ‘Social Assimilation Among American Catholic National-Origin Groups’, American Sociological Review, vol. 41, no. 6, pp. 1030-46.

BAUMANN, MARTIN 2000 Migration – Religion – Integration: Buddhistische Vietnamesen und

hinduistische Tamilen in Deutschland, Marburg: diagonal-Verlag. BERRY, J.W. 1980 ‘Acculturation as Varieties of Adaptation’, in Amado M. Padilla (ed.),

Acculturation: Theory, Models and Some New Findings, Boulder: Westview, pp. 9-25. BERRY, J.W. et al. 1989 ‘Acculturation Attitudes in Plural Societies’, Applied Psychology: An

International Review, vol. 38, pp. 185-206. CHÁVEZ, ALICIA FEDELINA A.F. and GUIDO-DIBRITO, FLORENCE 1999 ‘Racial and Ethnic

Identity and Development’, New Directions for Adult and Continuing Education, vol. 84, pp. 39-47.

CONSTANT, AMELIE, GATAULLINA, LILIYA and ZIMMERMANN, KLAUS F. 2006

‘Ethnosizing Immigrants’, IZA Discussion Paper No. 2040. DE DOMANICO, YOLANDA BAUTISTA, CRAWFORD, ISIAAH and DE WOLFE, ALAN S.

1994 ‘Ethnic identity and self-concept in Mexican-American adolescents: Is bicultural identity related to stress or better adjustment?’, Child & Youth Care Forum, vol. 23, no. 3, pp. 197-206.

GUISO, LUIGI, SAPIENZA, PAOLA and ZINGALES, LUIGI 2006 ‘Does Culture Affect Economic

Outcomes?’, Journal of Economic Perspectives, vol. 20, no. 2, pp. 23-48. HALLER, WILLIAM and LANDOLT, PATRICIA 2005 ‘The transnational dimensions of identity

formation: Adult children of immigrants in Miami’, Ethnic and Racial Studies, vol. 28, no. 6, pp. 1182-1214.

JACOBSEN, JESSICA 1997 ‘Religion and Ethnicity: Dual and alternative sources of identity among

young British Pakistanis’, Ethnic and Racial Studies, vol. 20, no. 2, pp. 238-56. KONG, LILY 2001 ‘Mapping “New” Geographies of Religion: Politics and Poetics in Modernity’,

Progress in Human Geography, vol. 25, no. 2, pp. 211-33. KURIEN, PREMA 2001 ‘Religion, ethnicity and politics: Hindu and Muslim Indian immigrants in the

United States, Ethnic and Racial Studies, vol. 24, no. 2, pp. 263-93. MCCLEARY, RACHEL M. and BARRO, ROBERT J. 2006 ‘Religion and Economy’, Journal of

Economic Perspectives, vol. 20, no. 2, pp. 49-72. NAGATA, JUDITH A. 1974 ‘What Is a Malay? Situational Selection of Ethnic Identity in a Plural

Society’, American Ethnologist, vol. 1, no. 2, pp. 331-50.

14

NAZROO, JAMES Y. and KARLSEN, SAFFRON 2003 ‘Patterns of identity among ethnic minority people: Diversity and commonality’, Ethnic and Racial Studies, vol. 26, no. 5, pp. 902-30.

ONISHI, AKIKO and MURPHY-SHIGEMATSU, STEPHEN 2003 ‘Identity Narratives of Muslim

Foreign Workers in Japan’, Journal of Community & Applied Social Psychology, vol. 13, pp. 224-39.

PARENTI, MICHAEL 1967 ‘Ethnic Politics and the Persistence of Ethnic Identification’, The

American Political Science Review, vol. 61, no. 3, pp. 717-26. RAJ, DHOOLEKA SARHADI 2000 ‘“Who the hell do you think you are?” Promoting religious

identity among young Hindus in Britain’, Ethnic and Racial Studies, vol. 23, no. 3, pp. 535-58. RELIGIONSWISSENSCHAFTLICHER MEDIENINFORMATIONSDIENST E. V., REMID 2006.

‘Religionen in Deutschland: Mitgliederzahlen.’ http://www.remid.de/ remid_info_zahlen.htm. SAEED, AMIR, BLAIN, NEIL and FORBES, DOUGLAS 1999 ‘New ethnic and national questions

in Scotland: post-British identities among Glasgow Pakistani teenagers’, Ethnic and Racial Studies, vol. 22, no. 5, pp. 821-44.

SOEP GROUP 2001 ‘The German Socio-Economic Panel (GSOEP) After More Than 15 Years –

Overview’, in Elke Holst, Dean R. Lillard and Thomas A. DiPrete (eds), Proceedings of the 2000 Fourth International Conference of German Socio-Economic Panel Study Users (GSOEP 2000), Vierteljahreshefte zur Wirtschaftsforschung (Quarterly Journal of Economic Research), vol. 70, pp. 7-14.

STATISTISCHES BUNDESAMT 2006 ‘Pressebroschüre “Leben in Deutschland – Haushalte,

Familien und Gesundheit, Ergebnisse des Mikrozensus 2005“’, www.destatis.de/presse/ deutsch/pk/2006/mikrozensus2005_b.htm.

15

Table 1. OLS of one-dimensional measurements of Muslim immigrants’ ethnic identity

Characteristics Language Culture Societal Interaction Migration Ethnic Self-

identification Ethnosizer

Age -0.012 (-0.51)

-0.045** (-2.29)

0.046** (2.36)

-0.049* (-1.72)

-0.018 (-0.68)

-0.016 (-1.12)

Age squared 0.0001 (0.21)

0.001* (1.93)

-0.001** (-2.50)

0.001* (1.67)

0.0003 (0.47)

0.0002 (0.77)

Age cubed -1.18E-08 (0.00)

-4.72E-06 (-1.49)

8.24E-06***(2.60)

-7.89E-06* (-1.69)

-1.60E-06 (-0.37)

-1.20E-06 (-0.53)

Age at entry 0.024*** (6.73)

0.007** (2.44)

0.001 (0.46)

0.005 (1.12)

0.003 (0.74)

0.008*** (3.82)

Age at entry squared -0.0002*** (-3.28)

-8.5E-05 (-1.54)

-8.36E-07 (-0.02)

9.31E-05 (1.14)

9.46E-05 (1.26)

-2.3E-05 (-0.59)

Female 0.064*** (3.02)

0.067*** (3.78)

-0.010 (-0.54)

0.013 (0.50)

0.049** (2.02)

0.037*** (2.88)

College or higher education in home country

-0.085 (-1.24)

-0.024 (-0.43)

0.039 (0.69)

0.022 (0.26)

0.018 (0.24)

-0.006 (-0.15)

Incomplete school education in home country

0.117*** (3.40)

0.062** (2.17)

0.125*** (4.35)

-0.005 (-0.11)

0.081** (2.07)

0.076*** (3.70)

Vocational education in home country

0.016 (0.46)

0.046 (1.60)

0.060** (2.08)

0.004 (0.10)

0.039 (1.00)

0.033 (1.60)

Complete school education in home country

0.060* (1.89)

0.060** (2.29)

0.051** (1.96)

0.005 (0.12)

0.081** (2.27)

0.051*** (2.73)

Ex-Yugoslavs -0.051* (-1.68)

-0.098*** (-3.84)

0.024 (0.94)

0.040 (1.06)

-0.100*** (-2.90)

-0.037** (-2.03)

Other ethnicities -0.124*** (-3.33)

-0.109*** (-3.54)

-0.151*** (-4.86)

0.079* (1.74)

-0.049 (-1.16)

-0.071*** (-3.19)

Constant 0.412 (1.22)

1.184*** (4.21)

0.068 (0.24)

0.780* (1.88)

0.853** (2.24)

0.659*** (3.28)

Adjusted R2 0.393 0.231 0.119 0.106 0.109 0.327

Number of observations: 430 Reference group: male, no education in the home country, Turkish Dependent variables: all six dependent variables range between zero (most commitment to Germany) and one (most commitment to the country of origin) * significant at 10% ** significant at 5% *** significant at 1% (two-tail test; t-values in parentheses)

16

Table 2. OLS of two-dimensional measurements of Muslim immigrants’ ethnic identity

Characteristics Integration Assimilation Separation Marginalization Age 0.161

(1.60) -0.089 (-0.90)

-0.002 (-0.01)

-0.071 (-0.70)

Age squared -0.003 (-1.46)

0.002 (1.01)

-0.0004 (-0.12)

0.001 (0.64)

Age cubed 2.15E-05 (1.31)

-1.8E-05 (-1.11)

5.63E-06 (0.25)

-9.39E-06 (-0.57)

Age at entry -0.025 (-1.60)

-0.050*** (-3.29)

0.079*** (3.71)

-0.004 (-0.25)

Age at entry squared -2.9E-05 (-0.10)

0.001** (2.18)

-0.001* (-1.75)

0.0001 (0.36)

Female -0.191** (-2.08)

-0.208** (-2.31)

0.304** (2.42)

0.095 (1.03)


0.118 (0.42)

-0.304 (-1.10)

0.280 (0.73)

-0.094 (-0.33)


-0.226 (-1.52)

-0.620*** (-4.28)

0.478** (2.37)

0.367** (2.47)

Vocational education in home country 0.022 (0.15)

-0.397*** (-2.73)

0.265 (1.30)

0.111 (0.74)


-0.030 (-0.22)

-0.427*** (-3.22)

0.567*** (3.06)

-0.110 (-0.81)

Ex-Yugoslavs 0.063 (0.48)

0.147 (1.14)

-0.419** (-2.34)

0.208 (1.58)

Other ethnicities 0.114 (0.71)

0.361** (2.30)

-0.365* (-1.67)

-0.109 (-0.68)

Constant -0.821 (-0.56)

3.069** (2.14)

0.885 (0.44)

1.867 (1.27)

Adjusted R2 0.115 0.211 0.242 0.031

Number of observations: 424 Reference group: male, no education in the home country, Turkish Dependent variables: all four dependent variables range between zero (not having any features attributed to members of this ethnic identity formation strategy) and five (having the features attributed to members of this ethnic identity formation strategy in all five factor groups) * significant at 10% ** significant at 5% *** significant at 1% (two-tail test; t-values in parentheses)

17

Table 3. OLS of linear measurements of Christian and non-denominational immigrants’ ethnic identity

Characteristics Language Culture Societal Interaction Migration Ethnic Self-

identification Ethnosizer

Age -0.011 (-0.64)

0.027** (2.30)

-0.008 (-0.59)

-0.019 (-0.99)

0.004 (0.18)

-0.002 (-0.15)

Age squared 0.0002 (0.69)

-0.001** (-2.19)

0.0001 (0.47)

0.001 (1.38)

-3.95E-06 (-0.01)

8.22E-05 (0.36)

Age cubed -2.05E-06 (-0.84)

3.34E-06**(1.99)

-8.31E-07 (-0.40)

-5.20E-06*(-1.86)

-8.52E-07 (-0.30)

-1.12E-06 (-0.72)

Age at entry 0.015*** (5.33)

0.006*** (3.30)

0.005** (2.00)

-0.001 (-0.44)

0.007** (1.99)

0.006*** (3.51)

Age at entry squared -6.8E-05 (-1.56)

-5.2E-05* (-1.72)

-2.1E-05 (-0.56)

9.44E-05* (1.89)

-1.7E-05 (-0.33)

-1.3E-05 (-0.45)

Female 0.020 (1.14)

0.011 (0.89)

-0.009 (-0.63)

-0.023 (-1.14)

-0.003 (-0.16)

-0.001 (-0.08)

Other Christians -0.019 (-0.73)

0.008 (0.46)

0.032 (1.46)

-0.011 (-0.37)

0.009 (0.31)

0.004 (0.24)

Non-denominational -0.022 (-0.70)

-0.014 (-0.63)

0.013 (0.50)

0.036 (1.02)

0.007 (0.20)

0.004 (0.22)


-0.073* (-1.66)

0.021 (0.69)

-0.055 (-1.49)

0.079 (1.58)

0.054 (1.06)

0.005 (0.19)


0.136*** (3.88)

0.088*** (3.66)

0.077*** (2.63)

0.043 (1.07)

0.110*** (2.72)

0.091*** (4.11)


-0.031 (-1.02)

0.012 (0.55)

-0.025 (-0.99)

-0.004 (-0.11)

0.012 (0.34)

-0.007 (-0.38)


0.095*** (3.17)

0.053** (2.55)

0.011 (0.42)

0.025 (0.72)

0.036 (1.02)

0.044** (2.30)

Turks 0.249*** (5.39)

0.203*** (6.38)

0.141*** (3.64)

0.050 (0.94)

0.103* (1.92)

0.149*** (5.10)

Ex-Yugoslavs 0.123*** (4.09)

0.065*** (3.16)

0.121*** (4.78)

0.084** (2.44)

0.013 (0.36)

0.081*** (4.26)

Greeks 0.191*** (4.74)

0.161*** (5.80)

0.109*** (3.22)

0.118** (2.57)

0.104** (2.24)

0.137*** (5.37)

Italians 0.194*** (6.40)

0.142*** (6.83)

0.123*** (4.85)

0.052 (1.51)

0.086** (2.45)

0.119*** (6.24)

Other ethnicities -0.016 (-0.51)

-0.028 (-1.29)

-0.047* (-1.76)

0.069* (1.89)

0.011 (0.29)

-0.002 (-0.12)

Constant 0.147 (0.59)

-0.153 (-0.89)

0.584*** (2.78)

0.355 (1.24)

0.279 (0.96)

0.242 (1.53)

Adjusted R2 0.280 0.228 0.148 0.045 0.043 0.197

Number of observations: 724 Reference group: male, no education in the home country, Catholic, Spanish Dependent variables: all six dependent variables range between zero (most commitment to Germany) and one (most commitment to the country of origin) * significant at 10% ** significant at 5% *** significant at 1% (two-tail test; t-values in parentheses)

18

Table 4. OLS of two-dimensional measurements of Christian and non-denominational immigrants’ ethnic identity

Characteristics Integration Assimilation Separation Marginalization Age 0.043

(0.62) -0.080 (-1.08)

-0.163* (-1.86)

0.201*** (3.26)

Age squared -0.001 (-0.35)

0.001 (0.88)

0.004** (2.09)

-0.005*** (-3.68)

Age cubed 1.65E-06 (0.16)

-7.94E-06 (-0.73)

-3E-05** (-2.38)

3.67E-05*** (4.10)

Age at entry -0.027** (-2.35)

-0.048*** (-3.77)

0.049*** (3.28)

0.026** (2.51)

Age at entry squared 0.0001 (0.77)

0.001*** (3.10)

-0.0003 (-1.37)

-0.0004*** (-2.65)

Female 0.095 (1.30)

0.003 (0.04)

-0.065 (-0.70)

-0.033 (-0.51)

Other Christians 0.101 (0.94)

-0.109 (-0.94)

0.034 (0.25)

-0.026 (-0.27)

Non-denominational 0.192 (1.48)

-0.098 (-0.70)

-0.138 (-0.83)

0.044 (0.38)


0.445** (2.49)

-0.309 (-1.59)

-0.027 (-0.12)

-0.109 (-0.68)


-0.571*** (-3.96)

-0.531*** (-3.39)

1.078*** (5.85)

0.024 (0.18)


0.110 (0.89)

-0.051 (-0.38)

0.068 (0.43)

-0.127 (-1.14)


-0.217* (-1.76)

-0.376*** (-2.80)

0.553*** (3.49)

0.040 (0.36)

Turks -0.089 (-0.46)

-0.631*** (-3.02)

0.794*** (3.23)

-0.074 (-0.43)

Ex-Yugoslavs -0.109 (-0.87)

-0.359*** (-2.65)

0.357** (2.24)

0.111 (0.99)

Greeks -0.349** (-2.11)

-0.616*** (-3.42)

0.799*** (3.77)

0.166 (1.12)

Italians -0.284** (-2.27)

-0.541*** (-3.99)

0.789*** (4.93)

0.036 (0.32)

Other ethnicities -0.088 (-0.67)

0.204 (1.43)

-0.172 (-1.02)

0.055 (0.47)

Constant 0.924 (0.90)

3.937*** (3.54)

2.312* (1.76)

-2.174** (-2.37)

Adjusted R2 0.113 0.160 0.206 0.054

Number of observations: 703 Reference group: male, no education in the home country, Catholic, Spanish Dependent variables: all four dependent variables range between zero (having no features which are attributed to members of this ethnic identity formation strategy) and five (having the features attributed to members of this ethnic identity formation strategy in all five factor groups) * significant at 10% ** significant at 5% *** significant at 1% (two-tail test; t-values in parentheses)

Documents

CLASH OF CULTURES: MUSLIMS AND CHRISTIANS IN THE